Cognee

What is Cognee?

Cognee is open-source software that gives AI agents a memory layer. It ingests your data, builds a knowledge graph using either generated or custom data models, then keeps it in vector, graph and relational stores so agents can recall connected context across sessions instead of only similar text chunks. It runs on Postgres with pgvector or on backends such as Neo4j, Kuzu, FalkorDB, Qdrant and Milvus, and connects to sources like Slack, Notion, Google Drive, GitHub, Confluence and Jira. Agents built on Google ADK, the OpenAI Agents SDK, CrewAI, n8n, LangGraph, Codex, Cursor or Claude Code can connect, and an MCP server covers other clients. You can self-host the SDK for free or use Cognee Cloud, which is multi-tenant with isolated workspaces. Cloud pricing is usage based: a Free plan with 1M tokens and one workspace, Standard at $1.00 per 1M tokens with extra workspaces at $5 per month, and a custom Enterprise plan delivered as a bring-your-own-cloud engagement.

Use cases

• Give an AI agent long-term memory
• Build GraphRAG pipelines
• Add semantic search over company data
• Index code repositories for agents
• Power customer support agents with context
• Run agent memory on existing Postgres
• Deploy memory in your own cloud

Who it's for

• AI engineers
• Agent developers
• Platform teams
• Enterprises

Why use Cognee

• Agents retrieve connected facts across sessions, not just chunks that look similar to a query.
• You can start free in the cloud or self-host the open-source engine with no license fee.
• It can add memory to a Postgres database you already run.
• Pricing is one flat rate per million tokens with no per-seat or per-document charge.
• Enterprise deployments can run in your own VPC with air-gapped options.

Cognee features

• Knowledge graph built from ingested data
• Combined vector, graph and relational storage
• MCP server for MCP-compatible clients
• Data source connectors for Slack, Notion, Linear and Google Drive
• Code indexing for repositories
• Multi-tenant workspaces with per-session memory
• Postgres with pgvector support
• Self-improving memory on Cognee Cloud
• Bi-temporal memory and provenance on Enterprise

Cognee integrations

• Claude Code
• Cursor
• Codex
• LangGraph
• n8n
• CrewAI
• OpenAI Agents SDK
• Google ADK
• Slack
• Notion
• Google Drive
• GitHub
• Confluence
• Jira
• Neo4j
• Qdrant

How to get started with Cognee

1. Create a free account at platform.cognee.ai, or install the open-source SDK to run locally.
2. Add data from files or connected sources such as Slack, Notion or Google Drive.
3. Cognee builds the knowledge graph and stores it across vector, graph and relational layers.
4. Your agent queries the memory through the SDK, API or MCP server to recall context.

Cognee limitations

• The Free cloud plan includes one workspace and 1M tokens
• Extra workspaces cost $5 per month each
• Cloud storage is capped per member regardless of plan
• Bi-temporal memory, provenance and BYOC are Enterprise features

Cognee pricing

Yes, Cognee has a free plan. Paid plans start at $1 usage-based, according to the official pricing page on 24 Oct 2026.

Cognee plans

• Free

  $0/month

  • 1M tokens included
  • 1 workspace
  • Unlimited users
  • Unlimited API calls
  • Agentic integrations (Claude Code, Codex, MCP)

• Standard

  $1.00/1M tokens

  usage based; extra workspaces $5 each per month

  • Everything in Free
  • Unlimited workspaces
  • Slack, Notion, Linear and Google Drive integrations
  • Code indexing for your repos
  • In-app support

• Enterprise

  Custom

  delivered as a BYOC engagement

  • Everything in Standard
  • Bi-temporal memory and conflict resolution
  • Provenance on every answer
  • BYO cloud supported
  • Dedicated support engineer and SLA

Plans as listed on the official pricing page on 24 Oct 2026. Prices can change and may vary by country or exclude tax. Check the official pricing page.

Cognee alternatives

The main Cognee alternatives listed on Debuly are Mem0, Zep, Letta, LangMem, Supermemory and ReMe. The table compares pricing and platforms; order is A–Z.

1. 

   LangMem

   Python library from LangChain that gives AI agents long-term memory and lets them refine behavior from past conversations.

   Developer platformsAutomation and agents
   Open sourceVisit LangMem (official site)
2. 

   Letta

   Platform, formerly MemGPT, for building stateful AI agents with persistent memory, from a terminal app, SDK or API.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Letta (official site)
3. 

   Mem0

   Memory layer for AI agents and apps, available as an open-source package or a managed platform that keeps user context across sessions.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Mem0 (official site)
4. 

   ReMe

   Open-source memory layer for AI agents that stores knowledge as local Markdown files with hybrid search and a linked graph.

   Developer platformsAutomation and agentsCoding
   Open sourceVisit ReMe (official site)
5. 

   Supermemory

   Memory and context API for AI agents with user profiles, a memory graph, retrieval and connectors, billed by usage.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Supermemory (official site)
6. 

   Zep

   Managed memory and context platform for AI agents, built on the Graphiti knowledge graph, with Python, TypeScript and Go SDKs.

   Developer platformsAutomation and agents
   FreemiumVisit Zep (official site)

Cognee FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from cognee.ai · fetched 23 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.